OBJECTIVES: To examine whether children compensate for participating in physically active behaviors by reducing activity at other times (the 'activitystat' hypothesis); or alternatively become more active at other times (activity synergy). METHODS: In 2002-2006, 345 British children (8-13 years) completed activity diaries and wore accelerometers. This generated 1077 days of data which we analyzed between-children (comparing all days) and within-child (comparing days from the same child). RESULTS: On week and weekend days, each extra 1% of time in PE/games, school breaks, school active travel, non-school active travel, structured sports and out-of-home play predicted a 0.21 to 0.60% increase in the proportion of the day in moderate-to-vigorous physical activity (MVPA). None of these behaviors showed evidence of reduced MVPA at other times, i.e. activity compensation (all p>0.15). Moreover, each 1% increase in weekday non-school active travel predicted 0.38% more time in MVPA at other times (95% CI 0.18, 0.58). This activity synergy reflected children using active travel for playing and visiting friends. CONCLUSIONS: Contrary to the 'activitystat' hypothesis, we found no evidence of activity compensation. This suggests that interventions increasing activity in specific behaviors may increase activity overall. The activity synergy of non-school active travel underlines the need for further research into this neglected behavior.